To refer initially only to my own country, the U.S. slid into a nuclear "weapons" option and system that is extremely and needlessly hazardous in several ways. However, the situation is redeemable. As Albert Einstein is quoted as saying, "The release of atomic power has changed everything except our way of thinking." But sometimes we can expand our way of thinking, even a way that has foolishly persisted since the 1940s.
How is this way of thinking hazardous? In several ways, but in part one we focus on one way, in part two on another. The danger of accidental war is heightened by the option of "launch of warning" and by the practice of bullying other powers with the bomb. Why? Because some warnings are false, and because, as Barbara Tuchman showed in The Guns of August, some wars start not by informed consent, but by an unintentional but unstoppable process.
These are two separate problems: the reliance on weapons many of which would be lost if not fired in response to an actual attack, and the dynamic of a nuclear "war" starting in an uncontrollable dynamic of threats and counter-threats among principals and allies. Let us look at each problem in turn.
We often try to understand new things in terms of what we've known. Thus, a computer screen is called a "desktop," and nuclear devices that can destroy most of a city in a flash are called "weapons." The digital desktop may be a harmless metaphor; the attempt to regard nuclear devices as weapons is not. Even when the discontinuity is actually extreme, his metaphoric habit persists.
To gauge how extreme, consider the London blitz of 1940-41. If Hitler had possessed even a primitive atomic bomb such as later destroyed Hiroshima, the blitz would have lasted one night and done enormously more damage. And it is often forgotten that hydrogen bombs can cause vastly more destruction than what was done in 1945.
As a thought-experiment, consider a single fictional device that could contaminate the whole globe with fatal radioactivity. Would we be right to call that a "weapon"? In Stanley Kubrick's movie, "Dr. Strangelove," the other side had built (but not yet announced) such a device. It was called a "doomsday machine." Actually, the thousands of nuclear devices that could be "delivered" in an "exchange," taken together, constitute such a doomsday machine, even without the amplification of the "nuclear winter" discovered by scientists in 1983.
The first point: though deployed by the military, nuclear devices are not weapons. In effect, they are a doomsday machine. The fact that we have survived this long is no comfort: one mistake would be fatal. At least one Secretary of Defense and one National Security Adviser have said publicly, when asked why we haven't suffered a nuclear exchange, "luck."
What is the second point? That the inability to protect land-based missiles in silos from nuclear blasts led to a dangerous option (or de facto policy) called "launch on warning." If the other side can disable enough of these missiles before our launch, we'd fail to "prevail." A problem has been that some of the warnings have been false, the falsity so far discovered in time. We don't know how many warnings have been false, but the documented cases are terrifying. For a detailed account see Eric Schlosser's's book.
Third point: the perceived need to launch land-based missiles before they are disabled by an attack means that a decision must be made by one person, the President, under his authority as "commander in chief" of our military. Say the perceived attack happens at night. This individual would have about 20 minutes, or even less in case of a "decapitation strike" directed against leaders (as from an enemy sub in the western Atlantic), to decide the fate of hundreds of millions of humans. He (or she) could "launch on warning under the rubric of "use 'em or lose 'em.""
The record of known errors is terrifying, yet we still keep "launch on warning" as an option. You have only to read Schlosser's New Yorker article (or, in my own case, have coffee in 1986 with an aide who was at Premier Khrushchev's elbow during the Cuban missile crisis).
Fourth point: you don't have to imagine a U.S. President who is impulsive or vengeful to see the folly of launch on warning. Many of the near-accidents occurred under leaders we have regarded as mentally stable, whatever our political views. Given our deployment of sitting-duck missiles, what is a President to do?
Could the U.S. take the lead by announcing that we will never launch on warning having rearranged our retaliatory forces so that they are invulnerable (as on subs) and thus don't ever need to be suddenly launched? If so, we could move toward a treaty providing verification that vulnerable weapons were, on more sides than one, being eliminated.